Assignment -
Part A - INTERACTIVE MODULE QUIZ
QUESTION 1 - 89.0% of the eligible subjects were interviewed for this study.
True
False
QUESTION 2 - What was the exposure and how was it assessed?
If they had smoked at any period in their life.
Age person started and stopped smoking.
Amount person smoked before onset of illness
The main changes in smoking history and the maximum ever smoked.
The proportion of pipes or cigarettes smoked.
All of the answer choices
QUESTION 3 - Who were the controls?
Patients with no cancer and patients with carcinoma of the stomach and colon.
Patients with carcinoma in any organ of the gastrointestinal system.
Patients with no cancer and patients with carcinoma of the stomach, colon and rectum.
Patients with carcinoma not associated with the gastrointestinal system.
QUESTION 4 - Based on the data in Table IV, compute odds ratios separately by gender.
Female OR: 2.5; Male OR: 14.0
Female OR: 14.0; Male OR 2.5
Female OR: 5.0; Male OR 7.0
Female OR: 7.0; Male OR: 5.0
QUESTION 5 - What type of data are in Figure 2?
Cohort
Cross sectional
Survey
Ecological
Cannot be determined from Figure 2.
QUESTION 6 - To adjust for confounders the authors matched both age and sex of cases to cancer controls.
True
False
Part B - Take Test: Homework - Case Control Studies
QUESTION 1 - Paragraph 3 of the Methods says that controls in six of the eight provinces were identified using provincial health insurance registration databases, while those in Ontario were identified using the Ministry of Finance Property Assessment Database, and those in Alberta and Newfoundland were selected using random-digit telephone dialing. What are strength and a limitation of each of these methods?
|
Health Insurance Database Strength
|
A. up-to-date census
|
|
Health Insurance Database Limitation
|
B. complete medical information
|
|
Property Assessment Database Strength
|
C. may be more subject to information bias
|
|
Property Assessment Database Limitation
|
D. an unbiased way of randomly selecting controls especially individuals with unlisted telephone numbers
|
|
Random-Digit Dialing Strength
|
E. restricted to persons with health insurance (selection bias)
|
|
Random-Digit Dialing Limitation
|
F. limited to property owners, excludes homeless and renters (selection bias)
|
QUESTION 2 - What type of cases were studied in the Kasim article?
Incident cases
Prevalent cases
QUESTION 3 - An investigator designs a case-control study of estrogen use and endometrial cancer. Cases are selected from his OB/GYN clinic located in a local hospital. This clinic is one of several OB/GYN clinics in hospitals servicing a large metropolitan area. Possible source(s) of controls include:
a. A random sample of women from the surrounding hospitals who have had a history of endometrial cancer, but are presently free from disease.
b. Women without endometrial cancer from the entire metropolitan area selected by random digit dialing.
c. Women at his hospital diagnosed with other conditions that are unrelated to endometrial cancer and estrogen use.
d. Women without endometrial cancer or any other type of cancer at his hospital.
e. B and C only.
f. B and D only.
QUESTION 4 - Case-control studies are most appropriate for
Rare exposures
Neither
Both
Rare diseases
QUESTION 5 - Calculate the odds ratio for the following 2x2 table and give its formal interpretation
|
Cancer (YES)
|
Cancer (NO)
|
|
High fiber diet (Yes)
|
475
|
790
|
|
High fiber diet (NO)
|
4411
|
7336
|
1.0 -Persons eating high fiber diets have the same odds of getting cancer as persons who do not eat a high fiber diets
0.5 - Persons eating high fiber diets have half the odds of getting cancer as persons who do not eat a high fiber diets
2.0 - Persons eating high fiber diets have twice the odds of getting cancer as persons who do not eat a high fiber diets
2.5 - Persons eating high fiber diets have two and a half times the odds of getting cancer as persons who do not eat a high fiber diets
QUESTION 6 - Because of cases deaths (15%), and physician refusal to give consent to contact some severely ill cases (8%), only 1545 cases were sent questionnaires. (Page 673, Methods Para. 2) What kind of potential bias does this fact create?
Selective survival
Unmasking bias
Diagnostic suspicion bias
Selective recall
QUESTION 7 - How might it alter the findings of the study if these people had been included?
a. It might have biased the results in a positive direction
b. The response rate might have been higher
c. Additional risk factors about adult leukemia might have been learnt because the people who died survived for less time and might have been more severely ill
d. It would have had no effect on the study
e. a & d
f. b & c
QUESTION 8 - Which of these methods do you think would produce the MOST relevant controls for this study?
Randomized digit dialing
Provincial health insurance registration database
Ministry of Finance Property Assessment Database
None of the above
QUESTION 9 - Using data in Table 1 (controls and all leukemia), calculate the odds ratio for female sex. Describe the result in a sentence.
OR=2.0-Women have twice the odds of getting leukemia compared to men
OR=1.00 -Women and men have equal odds of getting leukemia
OR=0.55-Women have half the odds of getting leukemia compared to men
OR=1.50-Women have one and one half times the odds of getting leukemia compared to men
QUESTION 10 - Can we conclude from Table 1 that occupational exposure to ionizing radiation is not a risk factor for leukemia? Why or why not?
No, OR=5.10 The odds of exposure in persons with leukemia is 5 times vs persons without leukemia
No, OR=14.66 The odds of exposure in persons with leukemia is 15 times vs persons without leukemia
No, OR=1.54 The odds of exposure in persons with leukemia is1.5 times the odds of exposure in persons without leukemia
No, OR=5.86 The odds of exposure in persons with leukemia is 6 times vs persons without leukemia
QUESTION 11 - Based on the data in Table 2 (years), is leukemia more strongly associated with residential or occupational exposure to ETS? Explain.
Yes, leukemia is more associated with occupational exposure >21 yrs OR =1.57, 95% CI (1.07-2.31). vs. residential 1.32 (0.88-1.95). The 95% CI for all levels excluding occupational exposure >21 yrs include 1.00
Yes, leukemia is more associated with occupational exposure >21 yrs because it has the highest OR =1.57
QUESTION 12 - In plain English what is the meaning of the odds ratios and confidence intervals for >21 years occupational exposure with ALL & CLL in Table 3?
There was no association of lifetime environmental tobacco smoke exposure among nonsmokers with acute lymphocytic leukemia (ALL). Nonsmokers with chronic lymphocytic leukemia (CLL) had a 2.37 times greater odds of having more than 21 years of occupational lifetime exposure to environmental tobacco smoke. Further they were 95% confident that the odds ratio of this exposure ranged from 1.30-4.32
None of the above
Nonsmokers with acute lymphocytic leukemia (ALL) had a 2.23 times greater odds of having more than 21 years of occupational lifetime exposure to environmental tobacco smoke. Nonsmokers with chronic lymphocytic leukemia (CLL) had a 2.37 times greater odds of having more than 21 years of occupational lifetime exposure to environmental tobacco smoke.
There was no association of lifetime environmental tobacco smoke exposure among nonsmokers with acute lymphocytic leukemia (ALL) or chronic lymphocytic leukemia (CLL).
Articles -
1. Smoking And Carcinoma Of The Lung: Preliminary Report, Author(s): Richard Doll and A. Bradford Hill
Source: The British Medical Journal, Vol. 2, No. 4682 (Sep. 30, 1950), pp. 739-748
Published by: BMJ Publishing Group
2. Environmental Tobacco Smoke and Risk of Adult Leukemia by Khaled Kasim, Patrick Levallois, Belkacem Abdous, Pierre Auger, Kenneth C. Johnson.
Attachment:- Assignment Files.rar